Metraipolis › archive › dc_895cea784a4b

KernelArc: A Multi-Agent Framework for GPU Kernel Optimization — arXiv:2608.17071

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.17071
License (as archived)
CC0-1.0
Archived
2026-10-18T14:28:29.612Z
SHA-256 of the archived text
c6f81b441ca3a8d99d4299cf625afa55712c327543d18f4e430cc62192a11f86
Stable URL of this page
https://metraipolis.com/documento/dc_895cea784a4b

How to cite

KernelArc: A Multi-Agent Framework for GPU Kernel Optimization — arXiv:2608.17071 — arXiv (en), https://arxiv.org/abs/2608.17071 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_895cea784a4b (Metraipolis archive dc_895cea784a4b, sha256 c6f81b441ca3a8d99d4299cf625afa55712c327543d18f4e430cc62192a11f86)
BibTeX
@misc{dc895cea784a4b,
  title = {KernelArc: A Multi-Agent Framework for GPU Kernel Optimization — arXiv:2608.17071},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.17071},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_895cea784a4b; sha256 c6f81b441ca3a8d99d4299cf625afa55712c327543d18f4e430cc62192a11f86},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.17071. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We present KernelArc, a multi-agent framework for autonomous GPU kernel optimization across heterogeneous workloads. Strategy-specialized agents run in parallel and coordinate through conclusions-only shared memory, a deterministic benchmark guard, and read-only cross-agent state with plateau-triggered drafting. We evaluate KernelArc on NVIDIA H100 and B200 GPUs using category-representative SOL-ExecBench workloads. The resulting implementations span custom BF16 GEMM, static cuBLASLt Expert-API configuration tables, fused mixture-of-experts backward, shape-gated decoder-layer fusion, native NVFP4 grouped-query attention, and paged prefill attention. In the public SOL-ExecBench leaderboard snapshot recorded on August~20, 2026, KernelArc ranked first on every representative L1, L2, Quantization, and FlashInfer task evaluated. The trajectories support the paper's central motivation: shared multi-agent search can broaden exploration and reach stronger incumbents within a fixed candidate budget, while the value of individual coordination features depends on the kernel and optimization stage.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-19 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Multiagent Systems (cs.MA); Performance (cs.PF).
Pagina del lavoro: https://arxiv.org/abs/2608.17071 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T14:27:54.235Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-19?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).